The objective of this research is to 1) identify and 2) describe the distribution of risk factors for cardiovascular disease (CVD) incidence and mortality in a cohort of free-living elderly persons. The identification of risk factors for CVD in older persons is important as 1) CVD is the leading cause of death and disability in elderly Americans and 2) there is evidence that CVD can be prevented or delayed in this age group. The data to be used in these analyses were collected in the Dunedin Program, a population-based geriatric health program designed to screen persons 65 years of age and older for a wide range of medical disorders. The Dunedin Program, located in Dunedin, Florida, has been in continuous operation since July 1975. Participants are screened annually, and extensive data on CVD and CVD risk factors are gathered during each visit. Information is available from questionnaires (i.e., family history, previous and present illnesses, drug use, smoking and alcohol use), physical examinations (i.e., EKGs, pulse rate, blood pressures) and laboratory measures (i.e., glucose, cholesterol, uric acid). Currently (1985), 5,085 elderly men and women have been screened at least once, and 1,540 persons have participated in the Program for a full eight years. This participation represents a total of 14,783 person-years of observation with an average follow-up of 4.9 years. The analysis will consist of three distinct phases. In the first phase, follow-up time for each participant will be computed, and changes in risk factor status will be analyzed. Descriptive information of the distribution of risk factors, and the prevalence of CVD in the entire cohort will be presented. CVD incidence and CVD mortality rates will be calculated for all participants. In the second analysis phase, incidence rates of CVD will be computed (number of events/person-years of follow-up) by category of risk factor level at baseline. The relative risk of CVD (incidence in exposed/incidence in unexposed) will be calculated for each hypothesized risk factor. The third phase will be multivariable analyses. Cox regression models will be used to determine independent and interactive effects of the identified risk factors on the incidence of and mortality from CVD in this cohort. This study will provide information on the prevalence and risk of CVD in a large, free-living elderly population. The potential identification of factors which may both increase the risk for CVD in older persons, and be modifiable or treatable is of significant public health importance, as modification of these factors may lead to a further reduction of events/deaths in this large and growing segment of our population.